Bitcoin Mining Companies Dump Over 19,000 BTC as Treasury Strategy Cracks
Public Bitcoin mining companies and corporate treasury holders sold more than 19,000 BTC in the span of a few weeks, forcing a hard reckoning with a narrative that looked bulletproof twelve months ago. Riot Platforms liquidated 3,778 BTC during Q1 2026 alone, selling at an average price of $76,626 per coin while producing just 1,473 BTC over the same period. The math is not complicated: these companies are selling inventory faster than they can mine it, and the market is noticing.
The Scale of What Just Happened
Start with the numbers, because they are brutal. Riot’s Q1 liquidation generated $289.5 million in net proceeds, but by the time the Q1 operational update landed on April 3, Bitcoin was trading near $66,867. That 12.7% gap between the average sale price and the current spot price tells you everything about how quickly sentiment has deteriorated. Riot ended the quarter holding 15,680 BTC, down 18% from the 19,223 coins it held a year prior. Arkham also flagged a separate 500 BTC outflow from a Riot-linked wallet on April 1, worth roughly $34 million at prevailing prices, a transfer Riot had not publicly commented on at time of writing.
MARA Holdings moved even faster. The miner sold 15,133 BTC between March 4 and March 25 for approximately $1.1 billion, using the proceeds to repurchase convertible notes due in 2030 and 2031 at an average discount of roughly 9% to par. That reduced outstanding convertible debt by about 30%, from $3.3 billion to $2.3 billion, saving an estimated $88.1 million in cash. CEO Fred Thiel framed this as deliberate capital allocation. Whether it was deliberate or forced is the more interesting question, and the company’s subsequent rounds of piecemeal layoffs across multiple departments suggest the answer leans toward the latter.
Then there is Genius Group. The Singapore-based AI education firm fully exited its Bitcoin treasury, selling the remaining 84.15 BTC in Q1 to retire an $8.5 million debt. At approximately $66,500 per coin at the time of liquidation, that sale crystallized a realized loss against an average acquisition cost basis of around $95,500 per coin. Genius Group had committed publicly to a Bitcoin-first treasury strategy in November 2024, pledging to allocate 90% or more of reserves to BTC. That pledge lasted roughly sixteen months. Nakamoto Holdings sold 284 BTC for about $20 million in March, representing around 5% of reserves, to cover working capital needs following acquisitions. Empery Digital sold 370 BTC at $66,632 to repay a term loan and unlock 1,800 BTC held as collateral. Combined with Riot’s Q1 activity, these companies alone account for a meaningful drawdown in corporate treasury supply.
Why the Selling Is Happening Now
The proximate causes vary by company, but the underlying pressure is uniform: Bitcoin fell from around $126,000 to roughly $66,000, a drop that transforms every treasury position opened above that level into an impairment liability. Industry analyst Kadan Stadelmann pointed specifically to the oil price surge connected to Middle East tensions that intensified in February as a primary contributor to elevated operational costs. “Bitcoin miners are being compelled to liquidate their holdings to maintain operational solvency,” Stadelmann said. Riot’s own data partially supports this: comprehensive power costs declined 21% year-over-year to 3.0 cents per kilowatt-hour, which sounds like good news until you consider that the company’s production fell 4% year-over-year to 1,473 BTC despite deploying 42.5 EH/s of hash rate, a 26% increase from the 33.7 EH/s it ran in Q1 2025.
That divergence between expanding capacity and shrinking output reflects the post-halving reality. More hash rate producing fewer coins at lower prices equals a revenue squeeze that no efficiency improvement fully compensates for. The network hash rate has actually pulled back from 1,160 EH/s to approximately 990 EH/s since early March, and mining difficulty dropped from 145 trillion to 133 trillion on March 20, according to Stadelmann’s analysis. At time of writing, the hash rate sits at 964.4 EH/s, with 106,523 blocks remaining until the next halving. The weaker operators are already shutting down, and what looks like a network stabilization is actually a culling process.
For the treasury-only firms, the mechanism is different but the outcome is the same. Genius Group lost access to capital markets after a U.S. court order temporarily barred it from raising funds and issuing shares. Without fresh capital, a Bitcoin-first strategy is simply a leveraged long position that has to be unwound when debt comes due. The strategy always required either rising prices or continuous access to equity markets. When both failed simultaneously, the exit was inevitable.
Bitcoin breaks below $60,000 before the end of Q2 2026 as continued forced selling from distressed mining and treasury firms overwhelms spot demand at current levels.
The Divergence That Defines This Market
While distressed sellers flooded the market, Metaplanet was buying. The Tokyo-listed investment firm acquired 5,075 BTC during Q1 2026 at an average price between $78,000 and $79,898 per coin, bringing total holdings to 40,177 BTC. That purchase pushed Metaplanet past MARA Holdings in the corporate ranking, placing it third globally behind Strategy and Twenty One Capital. Metaplanet’s CEO Simon Gerovich has framed Bitcoin as a long-term reserve asset suited to Japan’s inflation conditions and yen depreciation, and the firm has maintained a consistent accumulation pace since April 2024. The company is sitting on an unrealized loss of about 32%, given Bitcoin’s price near $66,400 at time of writing against an average cost basis of approximately $97,593 per BTC. Gerovich has signaled continued buying regardless.
This divergence is not random. It reflects two fundamentally different relationships with Bitcoin as a balance sheet asset. Metaplanet issues equity and runs options strategies against its holdings to fund accumulation, generating roughly 2.97 billion yen in Q1 2026 from those income strategies. It is structured to absorb drawdowns. MARA, Riot, and Genius Group were structured around the assumption of rising prices. When that assumption broke, the structure itself became the problem. The broader corporate treasury arms race that defined 2024 and 2025 has split into two distinct camps: capitalized long-term accumulators and distressed liquidators. The narrative that all corporate Bitcoin buying was structurally similar was always wrong.
Who Gets Hurt, Who Benefits, What Comes Next
The losers here are clear. Riot shareholders have already absorbed Wall Street’s verdict: Cantor Fitzgerald cut its price target to $29 from $31, Needham dropped to $24 from $30 citing mining underperformance and elevated costs, and H.C. Wainwright moved to $23 from $26. All maintained some form of buy or outperform recommendation, which reads as institutional inertia more than conviction. Empery Digital’s shares fell 75% from a 2025 high of $15.80. Genius Group’s equity is worth less than its remaining cash position after the Bitcoin exit. These are not paper losses waiting for a recovery. They are permanent capital impairments for investors who believed the treasury narrative.
The beneficiaries are more interesting. Metaplanet is acquiring scale at distressed prices while competitors shrink. Strategy, with over 762,000 BTC, has the balance sheet to absorb volatility that would destroy smaller players. The miners who survive the current culling, those with genuinely low power costs, diversified revenue from power credits and AI infrastructure, and manageable debt loads, will inherit network share from the operations that shut down. Riot’s $21 million in Q1 power credits, including $13.5 million from curtailment programs and $7.5 million through ERCOT and MISO demand response, hints at the operational moat that separates the survivors from the casualties.
The AI pivot that MARA is executing, and that Riot has telegraphed with its capital reallocation, is not a pivot born of strategic genius. It is a survival mechanism dressed in the language of vision. Selling Bitcoin to fund high-performance computing infrastructure is a bet that AI compute margins will compensate for the deterioration in mining economics. That bet might pay off. But the timing, forced by balance sheet pressure at the bottom of a drawdown, is not ideal, and the layoffs at MARA suggest the transition is costlier than the press releases implied. The structural shift among Bitcoin miners toward AI and debt financing has been building for months, but the Q1 data confirms it is now accelerating under duress rather than by design.
Public companies still hold approximately 1.165 million BTC across 195 firms, valued near $77.89 billion at current prices. That is more than 5% of the fixed 21 million supply. The selling wave of the past few weeks has trimmed corporate holdings by about 1%, but the structural pressure on the weaker holders has not resolved. More liquidations are probable as debt obligations mature, price support remains uncertain, and the firms least capable of accessing capital are the most exposed. The market is repricing what corporate Bitcoin treasury exposure actually means when the price stops going up.
Sentiment is the only asset class that moves faster than Bitcoin, and right now the sentiment around corporate treasury strategies is collapsing. The firms that built their identities around holding BTC are becoming sellers. That is not a technical signal or a macro thesis. It is a narrative inversion, and narrative inversions in crypto tend to overshoot badly in both directions. The companies still accumulating, Metaplanet chief among them, are making a long-term bet that this inversion is temporary. They might be right. But they are buying into a market where the sellers are not speculators trimming positions. They are operators keeping the lights on. That is a different kind of pressure, and it does not resolve on a quarterly earnings call.
